23 Apr
2005
23 Apr
'05
9:29 p.m.
Alan Taylor <alan-linuxaudiouser(a)lowb.org> writes:
On Sat, 2005-04-23 at 08:44 +0100, Rob Fell wrote:
When you double the sample rate you generally need to double the
buffer size. So, -r96000 -p1024 would be equivalent to -r48000 -p512.
Your latency is the same, either way.
--
joq
The most obvious reason I can see for this is
that the effective sample
rate is reduced to 12kHz via this method.
If you can increase your soundcard sample frequency, this should yield a
dramatic improvement (96kHz sample rate becomes effective 24kHz).
That makes sense, but I seem to be having trouble getting JACK to run at
anything above 48KHz. When I start it at 88.2KHz or 96KHz, I
immediately get messages such as "XRUN callback (xx skipped)", where xx
is a number between 50 and 300, depending on which sample rate and
frames/period I choose. I get at least 50 skipped callbacks per second
without doing anything but starting jackd.
If I am not mistaken, the Delta 44 should be able to handle at least up
to 96KHz. I believe my kernel is set up properly since I can record 6
or 7 tracks at 48KHz and 512 frames/period in RT mode for an extended
period of time without any xruns at all.
Any suggestions?